Hacker News Reader

Comment by SandB0x | Hacker News Reader

> The battle of devices has now become a war of ecosystems, where ecosystems include not only the hardware and software of the device, but developers, applications, ecommerce, advertising, search, social applications, location-based services, unified communications and many other things. Our competitors aren't taking our market share with devices; they are taking our market share with an entire ecosystem. This means we're going to have to decide how we either build, catalyse or join an ecosystem.

Please, please, Nokia, adopt Android and put your efforts into making great hardware to go with it, without ruining the interface like so many other manufacturers do.

View on HN

This is the way forward, a company that wants to develop native apps for mobile must already work with so many different systems. A new startup is going to focus on an iphone app, if they are successful or well capitalized from the start then they will look at android. The third choice in the market is shaping to be windows 7 phone.

Expecting companies to develop for then yet another platform, one which seems to be redefined often, is asking a lot and will result in getting only the really big players and lower rate developers trying to cash in on the under exploited platform.

In a way it's similar to Yahoo outsourcing their search to Bing, advertisers focused most of their attention on adwords and then also rolled some Bing campaigns, but only so much attention can then be given to the third player in the market for even less ROI for time spent due to the lower search volume.

Not really, Nokia's Ovi Store is now third in downloaded apps per day. However, in many countries such as in China, Ovi Store is the most popular app store.

http://www.forum.nokia.com/Distribute/Ovi_Store_statistics.x...

This says nothing of what the sales are like in China which is notorious for not paying for software. In China currently these phones have a much larger share of the market because they are cheaper than iphones and android phones.

Eventually when the prices of the better smartphones fall you would expect a big shift away from nokia unless they can improve on their current offerings. I think we will have to wait some time to for windows 7 phone to pick up some market share before app store figures show whether it is a better bet than nokia in the American market.

Pretty much the only people making money on Ovi are game companies. Nothing wrong with that, except that gaming is different to other types of software.

Also, note these points:

  * There are 30,000 apps available
  * 400,000 new Forum Nokia developers were added in the last 12 months.

That really illustrates what a horrible platform Nokia has. 400K new developers were interested in it over the last 12 months, and yet there are only 30K apps in the whole store. Compare that to the adoption that the iOS and Android app stores have had.

The prior development tools for native Symbian have been horrid, combined with difficult signing process etc. However, the transition to Qt and Qt Quick -based apps makes it all very easy and convenient thus enabling the potential better. It's true that it isn't the most exciting platform right now, but it's certainly on the right path.

Yeah, but that's like saying that Lotus 123 was on the right path when they finally ported it to Windows.

With Qt/Qt Quick you get an environment that is sort of competitive with where iOS and Android were over a year ago. Meanwhile, iOS and Android have moved on and are accelerating development of their SDKs, and Nokia is falling further & further behind.

Anyway, say you build your app, thinking how great it will be with the huge reach of the Symbian platform. You ship, and discover that (a) people who have a Symbian phone and can afford to buy your apps are just waiting to upgrade to iOS or Android, and so won't invest any time in new apps on their Nokia phones and (b) Symbian is only growing in use in the developing world, where people are very unlikely to buy your app because the cost is a lot more significant.

So you give up, and go back to iOS and Android where the toolsets are nicer and there is more money to be made.

Hmm, in my opinion Qt/Qt Quick is certainly superior to Android in many respects today. And I do use both.

> This is the way forward, a company that wants to develop native apps for mobile must already work with so many different systems. A new startup is going to focus on an iphone app, if they are successful or well capitalized from the start then they will look at android. The third choice in the market is shaping to be windows 7 phone.

I agree with what you said here. But to take it a step further I'd say that the number one rule I've learned from doing mobile native apps the last few years is that ideally one should NOT be making mobile native apps. Make a web app, that just happens to look and interact decently on a mobile device. Saves developers a lot of pain and a lot of time otherwise invested in a more narrow platform set of skills.

For this to be truly a reality you would need mobile developers to expose more functionality to the web app. The web app could also become a common interface with lower level functionality, so say a camera on an android phone and an iphone could both expose a common js interface.

This is what Palm was trying to achieve with WebOS.

> But to take it a step further I'd say that the number one rule I've learned from doing mobile native apps the last few years is that ideally one should NOT be making mobile native apps. Make a web app, that just happens to look and interact decently on a mobile device.

Yeah, make a web page so I would need to start the browser first, then wait for the page to load. The GUI and the experience will be completely different than anything what I already have on the phone, not to mention that the GUI will be suboptimal. Transition between screens in the app will be dependent on network coverage - if I'm in the bus and it stops on some blind spot just when I clicked the button, I'll have to watch white screen (or an error message) until the bus moves. And for what? To "save developers a lot of pain". Sounds reasonable, I'm sure I'll give you my money.

If your goal is to provide for a nice developer experience, sure. Thing is that developers are paid to do non-pleasant things in order to maximize the end user experience. Most web apps are crap on the desktop already, but not enough to be worse than desktop apps (because of other advantages); but on a mobile, they're 10x worse and in almost no cases equally good let alone better than the native app.

What web apps do you know that work great on a phone? Maybe my experience is skewed, but the slickest web-based phone app is the mobile version of Reddit, and that thing still sucks when you actually use it.

Color me unconvinced - native dev isn't going anywhere just yet.

WP7 isn't shaping up to be anything. Nokia sold more N8's than all the WP7 manufacturers combined.

Secondly, there is room in the market for other players as indicated by Bada's success.

I'd rather Nokia actually be the only truly open mobile platform out there. No, Android isn't sufficiently open.

It's my phone, at a minimum I should be able to pull all my own data off it in any form I choose.

IMO a GNU/free fork of Android would still be a better platform than MeeGo or Symbian, and would be cheaper to develop.

Yes, and you could put QT on top of it!

A C++ framework on top of a mobile OS in which most interesting things happen in the Dalvik VM? What for?

I think the hacker expectation is that a free open mobile OS would not be tied to a single programming environment and making it easy to port the mass of existing qt code would be useful.

Most interesting things including lawsuits. If things turn sour with Dalvik, Qt already runs on Android.

There is no such thing as a free Android, except if you tear the whole Java user space away.

And I'm REALLY curious about why you and 14 other people think that Android is such a good mobile OS. The VM is both a legal and speed liability, the Android UI is not GPU-accelerated and the entire OS is still immature (see the embarrassing bugs with the SMS or the browser security holes).

> There is no such thing as a free Android, except if you tear the whole Java user space away.

Isn't that part BSD-licensed?

> the Android UI is not GPU-accelerated

It's a bit hard when you have to support more than one device, but I believe it's a matter of time. As more and more hardware gains hardware acceleration, I assume it will come (if not already in the latest 2.x and 3.x series)

> the entire OS is still immature(see the embarrassing bugs with the SMS or the browser security holes)

Don't confuse the OS with the programs that run on it. You don't blame Windows when SharePoint eats up all your intranet, do you?

The license doesn't matter that much, it's the principles - one of them being that you don't get sued by such and such corporation if you use the code. Or Google doesn't change the terms if they lose on trial. No one has patents on C or C++ AFAIK...

SMS and web browser are core mobile OS services, they're not simple programs since connectivity is very important for a mobile phone.

> No one has patents on C or C++ AFAIK...

Oracle is not complaining Google is implementing Java in Dalvik. Oracle is complaining they are implementing things too similar to stuff Oracle has acquired patents about.

http://carlodaffara.conecta.it/oraclegoogle-the-patents-and-...

From a technical point of view the OS is actually the best out there, imo. It's developed from the beginning with handsets in mind and as such has quite some really nice features.

Which technical point of view? From an enterprise standpoint, where it's on par with the iPhone and behind Blackberry, or an API standpoint, where CocoaTouch has about 30 years of development lead-time?

How the point of being far older makes it technically superior is far beyond me. You wouldn't say Symbian is the best mobile OS out there because it's one of the oldest, would you? Also, afaik CocoaTouch has been developed for iphone/ipod and thus can't be 30 years old.

Anyway, features i find pretty neat are:

  - The concept of Activities
  - The concept of Intents and their connection to activities
  - real multitasking 
  - XML based layout with builtin internationalization
  - the use of Java as widely known language
  - Dalvik after getting JIT compilation = nice piece of technology, a fast and capable VM for such "slow" and limited hardware
  - oh and it's open source (except for the google apps, just like any other app)

Acitivites and Intents are just a great thing, love the concepts.

I’ll have to research Activities and Intents. It sounds counter to the app model on iOS, which—arguably—is bad for technical users. However, I consider the the fact that the UX enables my parents to use an app out of box, and not have to enable a DLL for some certain functionality. Either/or is sometimes an important part of UX.

I’m not sure why you like XML so much, is it that you can write an interface programmatically? While it’s true that a nib is nothing more than a writeout of a view’s object graph, it features a higher-level mode of internationalization (nib-level) meaning that interfaces don’t have to be shoe-horned into text fields. Besides, it’s always possible to create a blank view that’s drawn into.

However, Dalvik’s JIT is more a ripped feature of the JVM (hah), and Apple has worked against implementing similar, their memory management techniques are still mainly reference counting.

And the open source thing? Apple’s core OS is OSS as well, Google’s strategy is the same: open source the fundamentals, leave the important bits proprietary. Apple’s children are UX and native apps, Google’s is online apps, which is why the Open Source tide line is a bit lower on iOS: Google couldn’t care less about churn and boring, long-range updates, as long as they get the users doing web stuff. Both Google and Apple both protect what’s important to their business interests.

I didn't mean to start the whole apple vs google debate here..

But let me tell what i find so nice about activities and intents by example of a simple usecase. I am browsing in my RSS reader and find a good article. Now if i want to share this article, i usually have a "share" option which sends the intent to the OS. Depending on the registered applications it let's me send this article via mail or twitter or sms or even bluetooth (or really whatever app supports this intent). After deciding i want to send this as mail to a friend the mail activity opens up, let's me send it and seamlessly send me back to right where i was (reading the article). This is integrated throughout the whole OS.

Now, if i am reading some rss article in Pulse on my iPad, i have two options: Use the facebooks-share button of Pulse. Ok, but if i want to mail it or the app doesn't offer that at all? I'll have to copy the URL, close pulse, open the mail app, write the mail, close the mail app, open up pulse again.

Another example is: Google places offers me a restaurant i want to visit. Now depending on the apps i have installed i can use google maps, some other navigation software or the public transport app i have to show me the way to get there. It makes the user experience quite seamless, despite the applications not "knowing" of each other. I don't have to write down the adress, open up some other app, enter the adress again.

It's also the way to replace core functionalities of the UI. Want another home application? When pressing the home button on the device it's firing an intent and this can be handled by some home app of your choice. Same for the search button.

The nice thing about intents is that it's very flexible. Depending on the apps i have installed it offers me interaction between those apps.

About JIT: If you want to believe Wikipedia, JIT is no JVM originated feature but was invented somewhere in the 60s. I find it quite hard to imagine how the Dalvik JIT could be a real rip-off of the JVM JIT, as those are based on quite different architectures. The only thing in common is that they both use the same syntax.

About Open Source: I may be corrected here but it's perfectly possible to have a working phone by taking the sourcecode from android. It will have a running kernel, a working UI, i can install apps and live with it. What would be missing would be gmail or the android market. But there are other markets and other mail applications. Could you please point me to where i can find a compilable open source version of iOS?

The problem is that openness at the cost of user experience isn't good enough to keep a company alive. Ideally you'd get both, of course; but so far Nokia has been unable to deliver on user experience compared to iPhone and even Android.

Google is not delivering a truly open handset because it's in their corporate interest, not because it would detract from the user experience.

I'd rather wait for one company to try to and get their UX together than wait for Google to do an 180 on their data collection policies or for Apple to tear down the walls of their garden.

N900, the only phone you can own that you will really own yourself.

Which, coincidentally, can also run Android (if you really want to hack it that much).

...or MeeGo, and I guess Ubuntu and Debian are also possible. That is one part of the phone being truly open.

Here was the original announcement: http://flors.wordpress.com/2009/08/27/software-freedom-lover...

One of the big problems with Android at the moment is phone manufacturers not playing ball with its customers. Samsung have repeatedly neglected their phones in favour of releasing more new Android phones. At the moment, HTC is probably the king of Android phones, and most non-casual Android users will probably want to go with them, or with a newer hyped phone.

If Nokia were to enter the Android market and promote their open ethos throughout the Android community then the company could make a mint. Something as simple as releasing driver code for their phones so people can modify code to be run on their phones would put them miles in front with the Android community. If Nokia were to do these simple things then projects like the GAOSP wouldn't be so relied on to provide a usable phone with support.

Please stop thinking like a HN-er when discussing the entire market. If Nokia were to enter the Android market they would be just another Android producer. No one cares about drivers except hackers.

People don't care about different drivers, but they care when they want to download the Twitter app, only to realise that their new phone is on 1.6 and that they require 2.1, and Samsung will not upgrade their phone. A lot of open-source projects for community builds of Android get plenty of casual users wanting to get the most out of a phone they're currently tied to that the manufacturer has effectively dropped support for.

These problems exist because manufacturers have not thought about these issues, and would rather line their pockets with new customers than provide support for existing customers/suckers.

Frankly, I don't think most people care, either. They just want a phone that works, where they can find the features they use, and maybe some useful other things, like syncing contacts with gmail.

Most people I know with Android devices have never downloaded apps, some have downloaded a few in the first few days they had their phone, and only my most hard-core tech friends download any regularly. Most studies seem to confirm this behaviour. Most people don't even know what a 'phone update' is, and are caught off guard when their phone asks them about it.

I'd be interested in seeing these studies, as all I have to go on is personal experience and experience with the GAOSP, MustyMod and the popular custom builds on Android forums.

So, you buy a new Android phone, try to download the Twitter app, and realise that your 2 month old phone will never be able to run it. Does the phone work?

A lot of custom OS projects wouldn't exist today if consumers believed it did.

I googled a bit, but can't find any numbers any more. I thought I remembered a few 'studies' (I'm using this very lightly, this included self-reported sales numbers from manufacturers and blog posts from industry watchers with estimates) mentioned last fall. All I can find now is how many more average purchases iPhone users make (compared to Android), but that doesn't say anything about the shape of the curve of course - if iPhone users are overall equally likely to buy apps, and only 10% of the most hardcore Android users buy the same amount, the average for Android is down, but with a lot of users not buying (this is my assertion, like I said I can't back this up with credible numbers).

Anyway, my main point is that 'twitter' is something most users have only heard from on TV and don't even know what it is. For a regular user, a phone works when you can make phone calls with it, send text messages and set reminders/use it as an (alarm) clock. Then there are fancy features like using it for navigation (which is something they're already familiar with from standalone devices), contact/full agenda sync, and password management; and reading news sites; and post on facebook when they're so bored out of their mind that they really don't know what else to do.

And then there's the uber-advanced, only-nerds-do-this, full phone-centered lifestyle, where as much as possible is automated through phone apps. This is where twitter-client downloading people fall into - web nerds, turtle-neck wearing marketing people and art students. Of course in absolute numbers this is still a sizeable market, but it's only a fraction of the overall 'mobile phone' customer base. It will become more main stream in time, but not now...

How many OS projects are really used? A couple of hundred users, a couple of thousands for popular custom android roms? That's still marginal.

> phone manufacturers not playing ball with its customers

They do. Their customers are the telcos. most phone manufacturers will do whatever they are told, without much regard of how much this will affect the life of their users.

As long as we prefer to buy locked-up, subsidized phones from our telcos, this unfortunate arrangement will persist.

What about Meego? If they actually gave it some resources, rather than treat it as an afterthought, it could be great. To date Symbian has sucked all the resources from Meego, but one of the clearest messages was finally admitting reality on Symbian - 'if we continue like before, we will get further and further behind'.

War of ecosystems.. Google is their worst competitor. They either join smaller ones like WP7, WebOS, Bada etc. or keep on their own.

Reply on news.ycombinator.com